The US will create an AI system to diagnose diseases by voice
American scientists will develop software with artificial intelligence Voice as a Biomarker of Health, capable of analyzing patients of patients for diagnosis and studying diseases.
Experts from 12 research institutes, headed by the University of South Florida (USF) in compliance with confidentiality, will collect a training database of people. With the help of a dataset, they will train the AI system that doctors can use to identify potential diseases and neurological disorders by analyzing human speech.
The software will be trained to detect five types of violations:
- voice diseases – malignant and benign lesions of the larynx, paralysis of the vocal ligaments;
- Neurological and neurodegenerative disorders – Alzheimer’s disease, parkinson, stroke, lateral amyotrophic sclerosis;
- affective and mental disorders – depression, schizophrenia, bipolar disorder;
- respiratory diseases – pneumonia, chronic obstructive lung disease;
- Children’s voice disorders – speech or linguistic developmental delay, autism.
“These five Pharma Bro Martin categories of diseases have been selected on the basis of publications of scientists in the field of voice AI over the past 20 years,” said the head of the project and associate professor of the Department of Otolaryngology, Jael Bensussan in a conversation with The Register.
According to her, the voice is one of the simplest and cheapest biomarkers to study. The collection of such data does not carry physical risks for patients and can be carried out in conditions of limited resources, especially using modern technologies, added Bensussan.
To ensure confidentiality and security, experts use federal training with the support of AI Startap in the field of Owkin biomedicine.
US National US Health Institutes (NIH) will allocate about $ 14 million for a project for four years. In the case of the approval of the Congress, he will be able to receive additional funding.
In the first year, NIH will provide $ 3.8 million, which will go to collect a massive and diverse database of patients in clinical conditions. With its help, researchers will learn the general features in the sound of speech of people who diagnosed certain diseases.
Voice as a Biomarker of Health is part of the wider NIH initiative aimed at accelerating the introduction of artificial intelligence into healthcare. For four years, the organization has promised to invest up to $ 130 million in medical projects to collect data, developing tools and the creation of resources for the training of researchers.
Recall that in February 2022 Neuralink Ilona Mask was accused of cruelty to animals during testing neuroimtings.
In July, the Alphafold algorithm of the DeepMind and Laboratory has predicted the structure of more than 200 million proteins.
In June 2021, the WHO published the first global report on the use of AI in medicine and voiced the basic principles in its development.
Subscribe to FORKLOG news at Telegram: Forklog AI – all news from the world of AI!